Early Career Researcher Showcase

Centre for Languages, Culture and Communication

Gemma Ralton - Media framing of Generative AI across the UK Political Spectrum

Entry Number: CLCC 18

The release of ChatGPT in 2022 marked a turning point in the way we talk and think about AI. Yet little research has examined how generative AI specifically is framed in the UK media, which has important implications for regulation and democracy. Similarly, few studies consider the role of political ideology, which often shapes how people perceive and adopt new technologies.

------------------------------

Imperial Business School

Shutian Liu - When Nature Meets Social Class: How Genes and Parenting Reinforce Health Inequality

Entry Number: IBS 22

This paper examines how children's genetic predispositions influence parental health investments and, in turn, shape health outcomes such as obesity. Using data from the Millennium Cohort Study, the study applies family-based genetic designs to disentangle the direct effects of children’s genes from the indirect effects that operate through parenting behaviours. The analysis identifies three key genetic pathways: a direct influence of child genes on adolescent health, an indirect influence mediated by parental responses, and broader family-level genetic effects captured by parental genes. This work sheds light on how the interaction between nature and nurture contributes to early health inequalities across socioeconomic groups.

------------------------------

Department of Aeronautics

Charlie Aveline - Shape-changing mechanical metamaterials for cancer patient implants

Entry Number: AERO 1

Around 3.5 million people in the UK suffer from cancer with that number likely to grow. During stage 4 cancer, tumours can spread to other areas of the body, including the spine, causing severe pain and even fracture. Current treatments require highly invasive surgery which severely affects the patient’s quality of life in the terminal stage. 

------------------------------

Yaxin Mo - Reconstructing turbulent flows from sparse measurements with physics-constrained neural networks

Entry Number: AERO 11

Understanding flow behaviours, especially turbulent flows, is vital to many engineering applications such as drag reduction. Turbulent flow data are usually obtained through two ways --- simulations and experiments. However, both methods come with limitations. On one hand, simulations of turbulent flows to the required resolution are computationally expensive, sometimes prohibitively so. On the other hand, sparse measurements are often taken from flows during experiments. These measurements are usually a mix of pressure and velocity data and cover only limited areas of the flow domain. In this work, we develop a physics-constrained neural network to reconstruct the full flow field from such incomplete measurements. We demonstrate the capability of our network using both 2D and 3D homogenous turbulent flows. 

------------------------------

Flavio Savarino - How do shock-induced separated flows transition to turbulence?

Entry Number: AERO 12

The problem of laminar-turbulent transition of fluid flows has major implications in aeronautical engineering systems such as aircrafts, spacecrafts, aero-engines, to name a few. At supersonic speeds, the thin layer of fluid adjacent to the surface of a body may be disrupted by shock waves through a so-called shock wave-boundary layer interaction (SWBLI), which may lead to flow separation and turbulence. Since transition dramatically affects the aerothermal loading, understanding the underlying physical mechanisms is paramount to mitigate performance losses through flow control. Our work performs a theoretical and computational study on a SWBLI via a novel numerical framework that solves the governing Navier-Stokes equations in Fourier space. This enables accurate and interpretable prediction of the transitional mechanisms at a fraction of the computational cost of Direct Numerical Simulations, offering a tool to guide future control strategies aimed at exploiting the identified mechanisms for performance optimisation.

------------------------------

Maria Luisa Scarpa - Smart Control, Safer Robots: Differential Games in Action

Entry Number: AERO 13

Imagine a future where robots are dependable teammates—working alongside humans in factories, spacecraft, or hospitals. But in reality, many still falter in unpredictable situations. My research tackles this by using tools from game theory to improve how robots coordinate and make decisions. I design strategies that help robots plan, adapt, and collaborate more effectively—even in complex environments. For example, I’ve developed methods to simplify the control of multiple robots using differential games, with applications from spacecraft formation to energy-efficient teamwork. By bridging theory and practice, my work helps bring us closer to robots we can truly trust in the real world.

------------------------------

Ningyuan Fu - Shaping the Flow: Elastic Metamaterials for Drag Reduction

Entry Number: AERO 14

With the growing need for greener aviation, reducing skin friction drag on aircraft has become both an environmental and economic priority. One promising flow control method uses travelling surface waves to manipulate the turbulent boundary layer, which has been proven effective both numerically and experimentally. However, current techniques to generate surface waves face limitations on the frequency and wavenumber they can achieve, making them unsuitable for modern aircrafts. Thus, a novel approach of generating travelling surface waves via engineered metamaterials has been proposed. The metasurface is optimized to achieve the desired wave parameters and minimize energy loss. The final design will be tested in a wind tunnel experiment to explore its impact on the boundary layer at high Reynolds numbers for drag reduction purposes.

------------------------------

Department of Bioengineering

Ioana Esanu - The human body as a circuit board: or, what it takes to make a blood vessel

Entry Number: BIO 2

The human body is a complex machine, where specialised cells are constantly talking to each other. For example, when we exercise, our heart races and our blood flows faster. The blood vessels “hear” this signal and they widen, to let more blood through. My research focus is figuring out why in diseases like diabetes, this communication stops. My hypothesis is that it all stems from the patients’ red blood cells. Because this communication is still not fully understood, my work is to build models of blood vessels in the lab and simulate what happens in our bodies. By using simple materials like plastic and silicone and combining them with human cells, I can reproduce the flow of blood through our vessels and examine the effect of red blood cells. If my hypothesis is confirmed, I hope that my work will lead to new ways to predict and treat disease outcomes.

------------------------------

Pooja Kaur - Designed For Impact: Targeted Delivery of Cancer Therapies To Activate the Immune System

Entry Number: BIO 15

My research focuses on creating safer and more effective cancer treatments. I work with proteins called cytokines that help boost the body’s immune system to fight cancer. However, current cytokine treatments can spread throughout the whole body and cause harmful side effects.

To solve this problem, I am redesigning cytokines that can stick directly to the tumour site by modifying them so they recognise proteins that make up the tumour's structural support network. This helps keep the treatment focused where it’s needed. By doing this, we can activate immune cells to kill cancer and train the immune system to protect against future tumour regrowth.

This approach has shown promise in several types of cancers, including breast, colon, bladder and skin cancer mouse models. Excitingly, we’re seeing better treatment success, longer survival times, and in some cases, complete disappearance of the tumour — all with fewer side effects and stronger long-term protection.

------------------------------

Margherita Montavoci - Teaching the Immune System How to Fight Cancer

Entry Number: BIO 16

"Our own immune system can fight cancer, but sometimes it needs help to spot and attack the tumour cells. That’s what immunotherapy does: it helps the body fight cancer more effectively.

We are developing new drugs called “bispecific antibodies” that bind to two sites at once. One part binds to cancer cells. The other binds to special immune cells in our body called dendritic cells. These cells start the immune response and teach the body to remember the threat for future attacks. By linking cancer cells to dendritic cells, our antibody teaches the immune system to recognise and destroy tumours more effectively.

We are designing and testing different bispecific antibodies to see how well they activate the immune system. This strategy could lead to more precise and powerful cancer immunotherapy, especially for tumours that are difficult to treat or may relapse."

------------------------------

Amelia Clark - Fungi-on-a-Chip

Entry Number: BIO 17

Fungi are considered the hidden drivers of life beneath our feet, yet much about how they behave within soil is unknown. One reason for this is because it is hard to study fungi directly inside the soil itself. Devices containing microscopic channels in which single cells of fungi, called hyphae, can grow allow us to investigate how hyphae behave in an environment similar to the tiny pockets of air found in soil. Using these devices, my research focuses on how liquid is moved along the outside of hyphae and how this liquid is used by other microorganisms, such as bacteria, as a "fungal highway" to move from one place to another.

------------------------------

Department of Chemical Engineering

Suhaib Nisar - Optimising the fractionation of biomass with ionic liquids

Entry Number: CHEM ENG 19

Climate change and resource scarcity are driving demands in the chemical industry for sustainable alternatives to fossil resources, such as biomass feedstocks. Fractionation of lignocellulosic biomass, which makes up the woody parts of plants, is an important step for making available the biopolymers contained within, such as cellulose, lignin and hemicellulose. In this work, we modelled the fractionation of woody biomass through the ionoSolv process, a promising approach for obtaining fossil-free organic chemicals by employing low-cost ionic liquids (ILs). The multiple, partly unknown, underlying reactions and interactions complicate understanding the dynamics of the fractionation process. We developed a simple reaction network representing the bulk chemical transformations occurring. The trained models were exploited for a multi-criterion feasibility analysis to determine optimal operating regions for ionoSolv fractionation, which were then experimentally validated. Overall, this work provides a novel framework for the optimisation and prediction of ionic liquid-based biomass fractionation.

------------------------------

Robert Lever- Understanding the burden of producing therapeutic antibodies

Entry Number: CHEM ENG 20

Mammalian cells are used to produce many useful proteins, particularly antibodies for treating a range of diseases. To manufacture antibodies, cells are engineered to contain new genes to express antibodies. However, expression of new genes causes a burden, leading to cellular stresses, reduced growth, and lower yield. This is due to competition for a finite pool of cellular resources (ribosomes, polymerases etc.) known as resource competition. This research aims to understand this resource competition using a capacity monitor: a fluorescent protein whose fluorescence is proportional to the capacity of the cell i.e. the brighter the fluorescence, the greater the available resources. This tool helps to assess resource competition between different antibodies, to identify what makes an antibody burdensome. This can help to understand how to optimise design and expression of these antibodies for minimal burden: ultimately leading to easier, more robust, and potentially higher yield, antibody production.

------------------------------

Department of Civil and Environmental Engineering

Atif Rasheed - Shaking Less, Saving More: A Low-Cost Way to Stop Buildings from Collapsing in Earthquakes

Entry Number: CIV ENG 3

Earthquakes can cause devastating loss of life, especially when poorly built homes collapse in vulnerable communities. My research focuses on protecting such buildings—like mud or unreinforced masonry houses common in developing countries—by creating a simple, low-cost solution that reduces earthquake damage. Inspired by traditional Andean techniques, I use “Shicras”: bags made from strong rope and filled with stones and recycled rubber. These are placed beneath the building to absorb shaking and prevent collapse. Unlike expensive high-tech devices, this system uses affordable, locally available materials and can be installed without specialist tools. By allowing buildings to “shake less,” the Shicras reduce the forces during an earthquake while still safely carrying the building’s weight. This research bridges indigenous knowledge and modern engineering to create a protection system that is practical, scalable, and life-saving. It offers hope for communities that need earthquake safety the most—but can afford it the least.

------------------------------

Department of Mechanical Engineering

Yuankai Ren - Seeing Inside Batteries: Separating Structure and Material Changes Using Ultrasound

Entry Number: MECH 26

Batteries power many parts of our lives—from phones to electric vehicles—but what’s happening inside them is still a mystery in many ways. My research explores a new way to “listen” to batteries using harmless sound waves. Just like a doctor uses ultrasound to see inside the body, we use ultrasound to detect changes inside batteries—such as how their materials shift and how full they are—without opening them up. By studying how sound behaves inside different types of battery structures and materials, we’ve developed a general tool that helps researchers and manufacturers monitor battery health in real time. This method could make battery testing faster, safer, and more accurate, helping future technologies—from electric cars to smart grids—run more reliably and sustainably.

------------------------------

Catarina Duarte Nunes - What are my ammonia and hydrogen up to?

Entry Number: MECH 27

We cannot keep using diesel and petrol as fuels if we aim to save the environment. Ammonia and hydrogen are potential alternatives as neither contains carbon. To develop the new generation of engines which use these new fuels it is important to know how they behave and react. This work evaluates the accuracy of different mathematical equations to predict the behaviour of ammonia and hydrogen at different pressures and temperatures.

It was found hydrogen can be modelled using one of the simplest equations, the ideal gas law, with errors below 5% for density calculations when compared to experimental data, if pressures are kept below 100 atmospheres and temperatures above -70C. Ammonia however, requires a more complicated equation which is purposely fitted to experimental data to be accurately modelled, which is much more versatile than the ideal gas law and can be used at virtually any pressures and temperatures.

------------------------------

Natasha Waddell - Clear Coatings for the protection of historic ferrous objects stored in aggressive environments.

Entry Number: MECH 28

After an audit of English Heritage’s metalwork collection, it was found that ferrous objects were the most at risk of corrosion. With many of their sites of significant historical context, installation of environmental stability equipment is not always possible. Many of the sites are subject to fluctuating and aggressive environments, activating corrosion on ferrous objects. This research project aims to investigate and compare commonly used and industrial coatings protective clear coatings with the intention to tailor protocols for three EH case studies: J.W. Evans silver factory, Dover war time tunnels and Stott Park bobbin mill. The outcome intends to produce an evidence-based spectrum of clear coatings of a more sustainable production and safer use whilst broadening the palette for wider metalwork conservators’ community. This will be achieved through the combination of controlled lab-based accelerated ageing alongside on-site testing in order to produce realistic data on the protectivity of these coatings.

------------------------------

Department of Infectious Disease

Tochukwu Anyaduba - Reengineering Point-of-care Molecular Diagnostics for Cost Reduction in Sub-Saharan Africa

Entry Number: ID 4

Sub-Saharan Africa (SSA) harbours the highest burden of infectious diseases and antimicrobial resistance globally. Sadly, due to cost, many cutting-edge diagnostic technologies remain inaccessible in the region. This is particularly true for molecular diagnostics (mol.dx), which is expensive and unaffordable to the region except via foreign aids and subsidizations. The high cost of mol.dx, however, is partly due to its compensatory design strategy – a design approach that incorporates multiple processes and materials to address few limitations, inadvertently inflating costs.

Mol.dx is characterized by 3 basic requirements: cell lysis , cell-free nucleic acid capture and purification (CNAP), and amplification. We have identified a key design element of CNAP in state-of-the-art mol.dx platforms whose replacement reduces the materials, complexity, and cost of state-of-the-art mol.dx by ~70%.

By replacing this key element, our research presents advancements to mol.dx which include: Instrument-free mol.dx, ~70% reduction in platform cost, and reduction in medical plastic wastes. 

------------------------------

Samuel Channon-Wells - Beyond Symptoms: AI Maps Children's Diseases Through Blood Patterns

Entry Number: ID 23

When children are unwell with fever, doctors currently classify their disease based on symptoms and basic tests. But this approach often misses the molecular story unfolding inside their bodies.

My research aims to create a new disease classification system. Instead of looking at symptoms, we use artificial intelligence to analyse blood samples from nearly 5,000 children. By examining which genes are "switched on" in their blood cells we can define a molecular “fingerprint” for each illness.

Our analysis discovered three high-level groups: a healthy group, an antiviral group, and a bacterial/inflammatory group. Excitingly, within these groups we find distinct subgroups that don't match traditional diagnoses – potentially representing new disease entities. We're exploring the molecular drivers of these groups to identify treatment targets.

This classification system could revolutionize how we diagnose and treat childhood illnesses, ensuring children receive the right treatment faster and potentially revealing new treatment targets for poorly-understood conditions.

------------------------------

Department of Metabolism, Digestion and Reproduction

Tanweer Beleil - The Good, the Bad and the Ugly

Entry Number: MDR 29

Rupture of the fetal membranes (also known as when women break their waters) prior to 37 weeks of completed pregnancy and before the onset of labour is called PPROM (preterm prelabour rupture of membranes). If babies are born too early, they may have permanent medical problems. Pathogenic bacteria in the vagina increase risk of PPROM whereas dominance of the vagina by types of bacteria called Lactobacillus species, can offer protection against PPROM and preterm birth. This is because these types of “good” bacteria inhibit the growth of “bad” pathogenic bacteria through the secretion of antibacterial agents and the production of lactic acid. Understanding whether vaginal bacterial can reach the uterus to cause PPROM has important implications for the development of future targeted and preventative interventions aimed to selectively control pathogenic infection, leading to improved pregnancy outcomes.

------------------------------

Belen Gimeno Molina - Legally bound: the interaction case of immunoglobulins and vaginal bacteria

Entry Number: MDR 30

A healthy vagina is usually dominated by Lactobacillus species, while Lactobacillus absence is associated with inflammation, gynaecological conditions and pregnancy complications. However, it is unclear how vaginal bacteria is recognised and interacts with the local immune response. Here, I investigate how immunoglobulins can be bound or unbound to different bacteria, their association with the immune response, the binding point on the bacterial surface and their link with preterm birth (babies born too early, prior 37 weeks of gestation). In summary, IgG is bound to Lactobacillus crispatus, and immunoglobulin binding is associated with anti-inflammatory proteins and term deliveries. These results the development of immunotherapies for preterm birth prevention.

------------------------------

Jiwoo Choi - Next-Generation Shape and Stiffness Programmable Medical Devices: Thermally Drawn Shape-Memory Polymer Fibres in Medicine

Entry Number: MDR 31

Recent advances in fibre research have extended the use of fibres beyond just transmitting light, allowing them to become functional devices made from fibres. In this study, Shape-Memory Polymers (SMPs)—materials that can remember and return to specific shapes—are introduced into fibre manufacturing using a technique called preform-to-fibre thermal drawing. This new method creates Shape-Memory Polymer-based Fibres (SMPFs) with specifically designed cross-sections for each medical device. We demonstrate their practical uses, such as catheters with adjustable stiffness, neural implants that soften at body temperature, and cochlear implants whose shapes can be programmed. This research demonstrates the potential of SMP fibres in various advanced medical and commercial applications due to their adaptable mechanical properties.

------------------------------

Jennifer Fisher - Improving personalised care in inflammatory bowel disease

Entry Number: MDR 32

People with inflammatory bowel disease (IBD) have a higher chance of getting bowel cancer, which often starts with early changes in the cells called dysplasia. Our study looks at how removing these changes during a colonoscopy affects long term outcomes such as whether someone later develops cancer or needs surgery to remove their bowel (colectomy). We also tested the tissue around the area that was removed, even though it looked normal under the microscope. Using a new genetic test, we looked for small changes in the DNA called copy number alterations (CNAs). We found that if these changes are present, there may be a higher risk of developing cancer. This means that even tissue that looks normal could still carry hidden risks. Our research could help doctors decide which patients need closer follow-up and which ones are at lower risk, leading to more personalised care for people with IBD

------------------------------

National Heart and Lung Institute

Anies Makhdomi-Sohi - From Frankenstein's Monster to Cardiac Research

Entry Number: NHLI 6

The heart functions by making use of its own internal electrical system to bring about contraction. So, provided with the right nutrients, it will spontaneously beat. The technique, Langendorff perfusion, makes use of these principles to make an isolated heart function outside of the body. Langendorff perfusion allows one to study the heart and thus plays a pivotal role in understanding heart diseases.

In this study, we looked at the ability of the drug PT2567 to lower the risk of arrhythmia (irregular heart rhythm) in the disease Pulmonary Arterial Hypertension (PAH). PAH is a rare disease that causes high blood pressure in the lungs, which eventually spreads and increases the risk of arrhythmia. PT2567 was tested in rats with PAH and their hearts were studied via Langendorff perfusion. PT2567 had no effect on the diseased hearts, but the drug shortened the electrical makeup of hearts from healthy rats.

------------------------------

Aaraby Ragavan - Recovery from heart failure – is there such a thing?

Entry Number: NHLI 33

Dilated cardiomyopathy (DCM) is the most common heart muscle disorder worldwide and one of the leading causes of heart failure. A speedy diagnosis is key as lots of effective treatments exist, allowing clinicians to dramatically change the course of an individual’s disease. In some patients, we see their dilated hearts shrink down to a normal size, restoration of the heart’s ability to pump blood around the body and a total resolution of their symptoms. Our research focuses on these individuals who have recovered to try and understand why they recover and whether they will maintain this recovery. We will study their genes and medical test results to see how things evolve. We will scan the heart and learn how our heart muscle cells use energy and whether this differs between healthy individuals and those with recovered DCM.

------------------------------

School of Public Health

Pallavi Nair - Building Bridges in Dementia Care: Empowering Communities Through Human Connection and Innovation

Entry Number: SPH 7

Dementia is more than memory loss; it’s often fear, isolation and missed opportunities for early support. Across Northwest London, we are empowering communities by bridging care systems, cultures and conversation through a bottom-up approach, shaped by the community. Our work is interconnected through four community-led projects; i) ComPROACT, a co-designed training programme for Community Health and Wellbeing Workers (CHWWs) to identify early and provide culturally informed dementia care during their regular home visits; 36 CHWWs trained in three boroughs, ii)Exploring use of small devices (wearable technology) to monitor real-time changes in brain-body activity, iii)Dementia Tapestry, a partnership uniting healthcare providers, voluntary sectors and academia, iv) CommUniTEA creates safe spaces to open dialogue on brain health and dementia over tea. Every initiative is a bridge: From conversation to care, overlooked symptom to support and silo working to shared solutions. Together we are building dementia-capable communities where no one faces dementia alone.

------------------------------

Kimberley Foley - How can we use electronic health records to improve children’s health?

Entry Number: SPH 34

When you visit your doctor, have you noticed how much they type into their computer and wondered who, other than your doctor, might use this information? As a researcher, I use anonymous electronic health records to examine what happens during children’s visits to their doctor. I have used these records to describe the drop in the number of doctor’s visits by children during the COVID-19 pandemic, and the shift from face-to-face to telephone appointments. I have also described how patterns of doctor’s visits for respiratory infections changed after the pandemic ended, with a large increase in visits by children in December 2022. Next, I will look at what doctors do during these visits and describe the patterns of antibiotic prescriptions given to children. My research can be used by doctors and policy makers to help make sure the right healthcare services for children are available when they need it.

------------------------------

Jo Craddock - How adults born preterm view their health and what kind of support they need

Entry Number: SPH 35

Every year, about 1 in 10 babies worldwide are born too soon (before 37 weeks), and thanks to better care, most now survive into adulthood. But what happens after that?

Babies born early are often very small and sick at birth. Many need extra help to breathe, feed, and grow. While a lot is known about the challenges these babies face early in life, we know much less about what happens when they become adults. We know very little about what they think of their own health or what support they feel they need.

My research looks at these questions by reviewing past studies, running a UK-wide survey, and interviewing adults from many kinds of backgrounds. I want to understand what matters most to them, and what they think would improve healthcare for adults born preterm.

------------------------------

Husa Aldossary - When Good Recipes Turn Out Bad: Fidelity vs. Adaptation in Practice

Entry Number: SPH 36

Family recipes often lose their magic across generations. Beyond a decline in cooking skills, this might be due to modifications made by different individuals over time, to the point that the recipe no longer reflects the original. Modifications may be driven by several reasons, such as dietary needs, lack of ingredients, or time constraints. They can either improve the recipe or its acceptability or turn it into a cooking disaster.

In the same way, when a single health programme is modified in different contexts, depending on the modifications, it may drift from its original design, leading to less effective outcomes. Using the case of implementing the Brazilian-inspired Community Health and Wellbeing Workers programme in Westminster, my research explores how programmes evolve over time in real-world implementation, and how the tension between staying true to a programme’s design (maintaining fidelity) and making necessary modifications to respond to different factors is navigated.

------------------------------

Department of Surgery and Cancer

Ahmed Alboksmaty - "Hey Doc, Are You Listening?!”

Entry Number: S&C 8

Aren't you tired of competing every day for attention, not just with other people, but even with computers and screens? Imagine how much worse this feels in a medical consultation, where you’re trying to share your worries and concerns, but your doctor seems more focused on their screen, typing and taking notes, than on you. In my PhD, I’m introducing an idea to beat technology with technology! Clinical documentation is critical for our health, but what if we used AI-powered tools to automate these tasks? This would allow our clinicians and GPs to focus solely on us, listening, engaging, and communicating directly, without the distraction of typing. My research explores the impact of using AI-powered voice-to-text technology for clinical documentation in primary care. It’s a step towards reshaping primary care into what it was originally meant to be – a safe space for us to speak up and truly be heard.

------------------------------

Tricia Tay - Engaging older adults in digital interventions for healthy ageing

Entry Number: S&C 9

Have you ever wondered why it is so difficult to get older adults to pick up a new technology? How can we better utilise technology to promote healthy ageing for older adults? What is the impact of and engagement with digitally delivered interventions in healthy ageing? Through a systematic review, we aim to find out what digitally enabled interventions have been tested to tackle unhealthy ageing. My qualitative interview study highlighted reasons that stop older adults from using technology. We then co-design and implement studies to create lasting lifestyle changes to promote aspects of health in the form of an app-based intervention for older adults. These findings highlight the barriers and facilitators of engagement in digitally-enabled interventions for lasting behaviour change for healthy ageing among older adults.

------------------------------

Archana Gopalakrishnan - Active surveillance for prostate cancer : who and how?

Entry Number: S&C 37

Over 50’000 men are diagnosed with prostate cancer every year in the UK. Some of these cancers grow slowly and do not need treatment immediately. They can be monitored and treated if, and when they grow. This is called active surveillance. However, exactly who to include and the best way of following them up are still unclear.

Our trial, IP9-ATLAS, is comparing 2 approaches:

• the standard of care - blood tests every 3 months and a prostate scan only if we suspect that the cancer is growing,
• the newer approach - blood tests every 6 months and a prostate scan regularly, even if there are no signs of change.

Over 3 years, 1’263 patients will be included and monitored for 5 years, until 2032. The aim is to find out which method detects cancer progression earlier, so that patients can receive treatment when needed, with less cost. 

------------------------------

Jabed Ahmed - Improving patients comfort in colonoscopy at no extra cost!

Entry Number: S&C 38

Past experience or anticipated fear of pain and discomfort in colonoscopy is the main driving factor for patients not engaging in their appointments and follow up with colonoscopy. This can have a significant and detrimental knock-on effect in colorectal cancer diagnosis and delay or a missed opportunity in treatment of curable disease if picked up at a later stage.

There is a large amount of resource and investment placed on improving the endoscopy equipment with state-of-the-art material, improvement of video output and incorporation of artificial intelligence. This however will take time and financial investment and is very much a long-term strategy for the above problem.

The implementation of music and smartphone distraction techniques into the current colonoscopy procedure has the advantage of immediate capability of implementation with little difficulty and no extra resource acquisition. The result is an immediate and significant improvement of patient experience in a real-world clinical setting.

------------------------------

Parisa Naeli - Genetic speed bumps: when the code slows down, tiny proteins step in

Entry Number: S&C 39

Our cells rely on genetic messages to produce proteins in a process called translation. But not all messages are read equally. Some use “slow” codons—specific three-letter genetic codes—that slow down the translation, leading to reduced efficiency and message breakdown. We revealed how two proteins, ZC3H7A and ZC3H7B, help cells detect and manage slow messages. We found that ZC3H7A/B specifically bind to messages rich in slow codons and either block their translation or trigger degradation. They interact with stalled ribosomes, machines that make proteins; and recruit other cellular components to either stop protein production or remove the slow message entirely.

This process helps cells control which proteins are made and when, ensuring proper function and preventing errors that could contribute to disease, including cancer. The discovery highlights a new layer of gene regulation in human cells and may inform future approaches to targeting messages in medicine and biotechnology.

------------------------------

Emiko Sultana - Inequality of Pelvic Floor Surgery Access in the UK

Entry Number: S&C 40

Pelvic floor disorders like rectal prolapse, constipation, and faecal incontinence can severely affect quality of life, especially in women. Surgery is an effective treatment, however, not everyone in the UK has equal access to this care. This research will explore whether factors like income, ethnicity, or where someone lives affect their ability to receive treatment for these conditions. By analysing NHS data from the past decade, we will look for patterns in how long people wait, what types of procedures they receive, and how successful the outcomes are for different patient groups. We will also interview patients and healthcare professionals to understand personal and systemic barriers. We aim to highlight the areas where inequalities exist and suggest ways to make access to care available for everyone. This work supports the NHS aim to deliver high-quality, equitable, and patient-centred healthcare.

------------------------------

Department of Chemistry

Shradda Vadodaria - Drugging the “Undruggable” in Cancer

Entry Number: CHEMI 21

The term 'undruggable' was coined to describe proteins that could not be targeted pharmacologically. Proteins with vital roles in cancer promotion, such as Rab27a, have long been classified in this category due to their lack of well-defined binding pockets that can capture drug-like molecules. Our research directly challenges this paradigm with the discovery of a Rab27a Pocket that can engage small molecules with both high potency and selectivity. Rab27a has a pro-tumoral role in various cancers, including breast, melanoma, and brain tumors. Thus, inhibiting this protein provides a novel avenue for tumor suppression, presenting an untapped opportunity to target a viable and novel cancer target. Cancer is a highly evolving disease that requires sustained discovery of new targets and therapeutic modalities for its defeat by the drug discovery fraternity. Here, we present another prong in the fight against cancer immortality. Through innovative methods at the interface of chemistry and biology, we are pursuing the development of the first Rab27a inhibitors that can probe and disrupt cancer from the laboratory bench to the clinic.

------------------------------

Department of Mathematics

Jeffrey Tse - Is this really the cause? Instruments can help!

Entry Number: MATH 5

Learning about how one factor causes the other is of interest across all aspects of life. An example is the causal effect of obesity (e.g., Body Mass Index) on heart disease. A funnier question is: does ice-cream sale cause shark attacks? When experiments are not possible, we rely on instruments to estimate causal effects from observational data. An instrument is a special variable affecting the outcome only through the exposure. It is very powerful to remove bias. Without showing the mathematics, we demonstrate its beauty in two things: (1) how do instruments work? (2) what to do when not all instruments are true?

------------------------------

Ella Orme - How can we get the most information out of a study?

Entry Number: MATH 10

As technology has advanced, we now get lots of information about a patient in a cancer study. This can include details about their lifestyle from a questionnaire (e.g. whether they smoke), a snapshot of their DNA from a blood test or an image of their cancer from a scan. All this data can help us learn about the different types of a cancer that exist, and which treatment is best for which type. This knowledge can be used to develop tools to help a doctor choose the best treatment for a new patient. The problem is that all these pieces of information are very different and so we can’t just treat them as one big dataset. My work looks at how we can use these different types of data in the same tool, so that we extract the most information possible, maximising the valuable data voluntarily provided by cancer patients.

------------------------------

Timothy Kang - Signature Dimension Reduction

Entry Number: MATH 24

In most fields, datastreams are unquestionably ubiquitous.

A mathematical object called the signature is being used increasingly to analyse datastreams.

Intuitively, the signature of a datastream is a tool to encode all possible information about it. More mathematically, it can be thought of as an infinite-dimensional vector.

Since one cannot use infinite-dimensional vectors in practical computation, the truncated signature is considered. The truncated signature of a datastream can be thought of as a finite-dimensional vector which encode a certain level of information about the datastream.

However, truncated signatures can still be very high-dimensional, rendering their direct use computationally impractical.

In our current research, we map truncated signatures onto spaces of random matrices, which are lower-dimensional, and in certain conditions, preserve symmetries and patterns in the underlying data streams quite accurately. These maps would therefore provide a computationally efficient and feasible way to analyse many collections of datastreams reasonably accurately.

------------------------------

Lucas Moschen - Steering Crowd-like Systems: An Optimal Control Approach

Entry Number: MATH 25

Pedestrians crossing a plaza, opinions spreading through a community, and molecules drifting in a liquid all share a common feature: they behave like countless particles moving together, making it impractical to track each one individually. To manage this complexity, we treat these systems as a single flowing "crowd", described by its overall density over time rather than by following each component (a so-called "mean-field" approach). This project aims to steer the crowd's density towards a desired state by minimising the external force needed to reshape the density. Essentially, we compute the smallest possible "push" required to guide how the entire group moves. By applying optimal control techniques, we can drive the distribution as intended. Although our work is theoretical, the methods we develop could help in practical scenarios such as planning efficient evacuations, accelerating simulations in chemistry and physics, or influencing how ideas spread online.

------------------------------

• The Showcase is open to the following groups:
  
  • Imperial College Postgraduate Doctoral Students
  • Imperial College Research Staff
• All participants are required to present their research at the event on Wednesday 02 July 2025 from 11:00-13:00 in Seminar Rooms 119-122 in the Sir Alexander Fleming Building.
• All participants will be allocated a 1m x 1m poster board or a table (to be chosen upon application).
• Participants can use their poster board or table to display anything that will assist them in communicating their research. This could be a research poster, a selection of images/artwork or any other visual aids/props; this is your chance to be creative!
• A £25 per participant expenses fund can be claimed to support the purchase of any materials or printing costs. (full details of how to claim to be provided)
• Participants will be assessed by a selection of judges which will include Imperial staff and students from all levels of study.
• Participants will be asked to present their research individually to each judge; the pitch should not exceed 3 minutes and must be aimed at a layperson.
• Participants will be judged on their ability to explain their research in a way that can be shared with the general public and on how well they visually & creatively communicate their research (full judging criteria detailed below).

**Applications are now closed**

The following prizes will be awarded:

• 1st Prize - £500
• 2nd Prize - £250
• 3rd Prize - £150
• People's Choice - £100 (voted for by visitors to the event)
• Creativity Prize x 4 - £50 (awarded to our most creative participants)